New Attacks Trick OpenClaw AI Agent Into Running Code and Leaking Secrets

Multiple high-impact developments: researchers demonstrated attacks that force OpenClaw AI agents to execute attacker-controlled code and exfiltrate secrets; a new BitLocker bypass dubbed GreatXML was disclosed; a financially motivated ransomware operation (“The Gentlemen”) claims widespread victims and worm-like spreading; and researchers published a proof-of-concept self‑replicating AI worm using local open‑weight models. Concurrently, vendors pushed urgent patches and fixes — including a record Microsoft update, active exploitation of a Chrome V8 zero-day, and several critical RCE and info‑